Hi all! First, thanks to David Kimdon for pointing out the docs, I tried as best as I could to fix it but alas! I gave it up in the end.
Now I�ve restored the MBR with a DOS floppy/fdisk, and boot linux from a floppy. (Thanks to whoever put the "fdisk /mbr" tip in the man page for grub!) I�m beginning to believe that no-one ever really put much effort into writing a bootloader that can handle modern hard disks. I�ll just stick to my boot floppy (or maybe try the entirely different alternative that�s available to us who, for one reason or another, hold on to MS: loadlin) until someone does. (No, I won�t volunteer to do it unless I win a lottery and get rich - I work with programming all day and when I get home, more programming is not exactly what I long for...) Keep on hacking out there! /Christer > -----Original Message----- > From: Eduard Bloch [mailto:[EMAIL PROTECTED]] > Sent: den 4 oktober 2002 13:26 > To: Christer Jansson A (QRA) > Cc: '[EMAIL PROTECTED]' > Subject: Re: LILO fails > > > #include <hallo.h> > Christer Jansson A (QRA) wrote on Mon Sep 30, 2002 um 02:15:55PM: > > > LILO prints "L", followed by the number "07" which seems to > repeat endlessly. > > I think, D-I will face the same problem with any boot loader, > sooner or later. > If the disk presentation of the kernel differs from that used > by BIOS/DOS, > even Grub is doomed to fail. > > Is there any effort to make Grub/Lilo/Syslinux to tell its > (read: BIOS) view > of the disks/geometries to the kernel, so we can use this in > D-I to make the > boot-loaders configuration right? Imagine, we have Grub as > loader for D-I. On > boot, it gives the kernel a parameter like > hdconf=0x80=serial-nr,0x81=serial-nr,... etc, so we can scan > the devices with > hdparm later and compare them with hdparm's output (or > something similar for > scsi). > > Gruss/Regards, > Eduard. > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
